HADV schemes in CCTM

CMAQ
1

Hi,

When reading CMAQ User’s Guide, I find that there are three optional schemes to solve the PDE in HADV. They are: YAM, BOT and PPM. However, in the version 5.3.1 and hereafter, only PPM is included in the source codes.

My question is, why the other two schemes are excluded? Is it because that they are incompatible to the new versions, or that PPM is the most desirable scheme and the other two are not recommendable?

Cheers,
Highray

2

@highrayzh,

Please note a couple of things:

  1. The YAM scheme isn’t actually the scheme developed by Yamartino. So it’s a bit of a misnomer. In fact that scheme is just PPM with minor edits to improve cross term errors when alternative X and Y advection.

  2. The BOT scheme isn’t fully implemented in previous CMAQ versions, so be careful using code that may provide that option. Check to see if the scheme is fully implemented.

  3. The PPM scheme is very-widely used for advection in AQ models and has several desirable characteristics that make it ideal for gas dynamics. For CMAQ, this scheme is our recommended scheme based on extensive testing and comparison with other schemes such as BOT for AQ applications.

Hopefully that clears things up for you!

2 Likes
3

Hi Fsidi,

Thanks a lot for your detailed and nice reply!

I still have a few small questions on the document I am reading (Microsoft Word - ch7_final (simsage.co.uk).

  1. According to the tables in the document, YAM is much better than PPM in those cases. Is this YAM you replied to me as a misnomer (actually PPM with minor edits), or the real Yamartino’s scheme?

    fig.1
    fig.11377×403 35.9 KB

  2. I find that in the recent releases of CMAQ, only PPM is included. Does that mean that the YAM scheme (namely, PPM with the minor edits) is not as desirable as PPM? Or, the effect of such edits turned out to be a little unsatisfied?

Cheers,
Highray

4

@highrayzh, Great questions!

  1. The YAM in this paper is the actual scheme developed by Yamartino (1993). The predecessor to CMAQ (MAQSIP) and initial versions of CMAQ had multiple advection schemes as noted in that paper. Each of these schemes was tested thoroughly using idealized and real test cases. The PPM scheme was more frequently used because of monotonicity (unlike YAM and BOT), positive-definite characteristics, mass-conserving characteristics, use with irregular grid spacing (unlike say ASD), and computational performance across a variety of tests (idealized and 3D model based). Over time, other schemes were dropped because of maintenance costs for little-to-none used options.

  2. The “yamo” scheme misnomer, came to be about when Bob Yamartino proposed the minor edits to PPM to the EPA. The edits themselves were causing various issues so starting v5.3, we decided the costs outweigh the benefits. For the official release note see CMAQ/CMAQv5.3_PPM.md at main · USEPA/CMAQ · GitHub.

2 Likes
5

@fsidi ,

Thank you sincerely for your clear and great replies!

With your help, I finally figure out the relationships between these schemes.

Have a nice weekend!

Highray

1 Like